WORK-ZONE SAFETY ITS Technical Analysis and
Contents
1. relative to the passive technologies and the challenge of running them at high enough burst rates to detect fast moving vehicles 3 5 MAGNETIC Magnetic sensors use a dual axis flux gate magnetometer to measure the Earth s magnetic field When a vehicle approaches the detector the vehicle distorts the magnetic field and the sensor detects this change Both the SPVD 2 Self Powered Vehicle Detector Midian Electronics and the Groundhog Permanent Count Station Nu Metrics use this method They are packed in plastic canisters and buried in the roadway The SPVD 2 is powered by a 13 5 Volt 17 amp hour alkaline D cell pack that can last 4 to 5 years The sensor sends vehicle arrival and departure messages to an above ground receiver via a 47 MHz FM radio modem This sensor technology is passive and consumes very little power which satisfies one of the primary requirements in the smart barrel application Banner Engineering Corp has recently introduced the M Gage S18M Vehicle Detection Sensor which sells for 209 in quantities of one This product uses a 3 axis magnetometer and is targeted for both above and below grade installations Background condition and sensitivity adjustments allow tailoring the sensor to the applications magnetic environment object properties and desired range Several sensor manufactures Honeywell Crossbow and Fraunhofer Institut Photonsche Mikrosysteme sell two and three axis magnetometers As discussed in the2. 18 Distribution Statement work zone collision avoidance ITS rear end Unlimited collision speed sensor sensor technology 19 Security Classification of this report 20 Security Classification of this page 21 No of Pages 22 Price None None 22 il TABLE OF CONTENTS Pe SIAN OU 1 O Mice Sn aan an ence asst 1 2 Concept of the Work zone Safety ITS System 2 21 System Cono mr aO Sen me tt iranien 2 SMA BANE RE a T A 2 BAT SUD VISOR ised ie E ea 4 SE OUDOLY ISO Ana ann ae E ee nn 4 22 Setup and C ANDTATIOM 2 Een ni 5 DST AL OM een E S 6 Determining signal intensity based on distributed speed measurements 6 Data collection and POCESSIDS entend a 6 3 Review of Detection Technologies Products 11 3 PRC AVS Mre o D uen 11 32 Passive NTM SM drae 11 3 Passive ACOUSTIC cuinen mr a 12 3A UASO tn ed E OE O ei tie 12 JI Mae e ea A T T 12 3 0 MITON AVE a a ean eanaeeeetisod ease 13 IA NV A Ona E E a te RE 13 4 Selected Technologies for Evaluation Vs 14 AL Dual Passive IN tare sess ne NN ne 14 42 Dual zaxis Mapnetomete roserne upnene 14 AS DualActive Maei nd outa vostenntelaeuatteesniceniaeas 15 4A Pulsed Doppler UIASOMNdiss nine nee dents 15 45 Dual UitraSOniG Presence dansant 16 5 Sensor Evaluation and testing essaient auraient 17 Gi BIBHOS DANSE ne niet os 18 iil 1 INTRODUCTION This document is an interim report by the University of Michigan Transportation Research Institute to the Fe
3. WORK ZONE SAFETY ITS Technical Analysis and System Proposal Michael Hagan Christopher Winkler June 2004 fa ING ih a A Yee pay wie lite F i f Transportation Research Institute WORK ZONE SAFETY ITS TECHNICAL ANALYSIS AND SYTEM PROPOSAL Michael Hagan Christopher Winkler The University of Michigan Transportation Research Institute Ann Arbor Michigan 48109 2150 U S A Report No UMTRI 2004 xx for Federal Highway Administration Contract DTH61 01 C 00049 June 2004 Technical Report Documentation Page 1 Report No 2 Government Accession No 3 Recipient s Catalog No UMTRI 2004 15 5 Report Date 4 Title and Subtitle J une 2004 Work Zone Safety ITS j 6 Performing Organization Code Technical Analysis And System Proposal 7 Author s 8 Performing Organization Report No Hagan M and Winkler C 9 Performing Organization Name and Address 10 Work Unit no TRAIS The University of Michigan Transportation Research Institute 11 Contracts or Grant No 2901 Baxter Road Ann Arbor MI 48109 2150 U S A 12 Sponsoring Agency Name and Address 13 Type of Report and Period Covered Federal Highway Administration Interim report U S Department of Transportation 14 Sponsoring Agency Code 400 7th Street S W FHWA Washington D C 20590 15 Supplementary Notes 16 Abstract This interim report presents the broad concept for an adaptive tra
4. ast the sensors in accordance to a matrix of prescribed speeds and ranges Target vehicle speed will be measured independently with conventional radar gun Within practical limits tests will be conducted under varying environmental conditions as appropriate for challenging the particular sensor Test sites will be limited to UMTRI or University property and local roadways The data records of these evaluation tests will be examined and analyzed to choose the set of the two or three most promising sensors These sensors will then be deployed in a limited field test wherein the sensors will operate unaccompanied for periods of several days at each of at least two road sites At least one site will be characterized by repeated queue and by the expectation of variable speeds at least through the mid to slow speed range Such a site might be an actual work zone but could also be near a stop light on a nominally fast moving urban arterial At least one other site which will insure high volume high speed traffic will be employed e g a nearby limited access roadway The sensors will be installed in one or perhaps two conventional traffic control barrels Data again will be gathered using existing UMTRI DAS In addition to the sensor data we will collect reference vehicle speeds using a conventional radar speed detector and a continuous video record using video camera and the frame grabber video storage feature of the UMTRI DAS As an additional source
5. ated intensity would be adjusted either proportionately or in steps to higher levels up to a maximum intensity associated with an upper deceleration threshold Also the differential speed of the vehicle above the posted speed Le Si Sposted Would be compared to a another set of thresholds to determine an intensity setting based on over speed The maximum of two intensities so determined would actually be used For barrels where the intensity is not calculated by the preceding procedure 1 e the positions i 1 where si is not current the signal intensity would be set equal to that of the preceding barrel The process would be progressive such that a currently calculated intensity would propagate down the line until reaching a position where a different intensity had been calculated 1 e a position where a different vehicle established the rate Finally where it may be advantageous for paired sets of speed sensor and signaling device to be widely spaced the smart barrel concept would require slight modification Namely where spacing is so wide that the next barrel is not readily visible the signal device must be spaced down stream from its associated speed sensor in order that the adaptive signal calculated on the measured speed could actually be displayed to an isolated vehicle passing the station Thus a widely spaced barrel would actually have to be a pair of barrels or perhaps a single barrel followed by an associated barrel
6. deral Highway Administration under Contract DTH61 01 C 00049 Work Zone Safety ITS Estimates of the increased crash risk in work zones have varied widely from 26 to 168 depending on the circumstances Despite the lack of quantitative precision it is evident that work zones increase crash risk The main objective of the Work Zone Safety ITS project is to develop a technology that directly reduces the incidence of work zone crashes Contributing factors cited in work zone crashes include the speed differential between two vehicles irregular maneuvers and speeding in challenging locations Of these speed differential is the most obvious explanation for the prevalence of rear end collisions in work zones A simple theory of work zone risk suggests that as traffic density in a work zone increases traffic slows down in the heavily congested areas causing a backup of slow moving traffic upstream of the congested area The speed differential between slow moving traffic and the traffic moving at posted speeds becomes a hazard because of the driver s uncertainty about where it occurs the magnitude of the speed differential and the span of roadway over which the differential is observed This document first presents the broad concept for an adaptive traffic signaling system for work zones based on a distributed system of traffic speed sensors and traffic signaling devices Both the sensors and the signals are seen primarily as rather simple inexp
7. ecelerations decji r can be calculated according to 1 1 J Adi Siting for all j gt 1 and where decyjj j 1s the deceleration in gravitational units g djirj 1s constant predetermined at the time of set up and is the distance between barrels 1 and j in meters Si Sj are the speeds at barrels 1 and jJ in meters second tlag is a constant representing the total time lag system latency and driver reaction in seconds grdi ei e djir predetermined at the time of setup is the average downgrade between barrels i and J ei are the elevations at barrels 1 and j in meters and 0 102 is the conversion constant for deceleration in g per m s Note that in equation 1 the first term on the right side is the actual required deceleration and the second term on the right side is an adjustment for average road grade increasing the apparent acceleration demand on downgrades and decrease it on upgrades thus providing the grade compensation discussed previously From these values the required deceleration for the vehicle currently passing barrel 1 decreq i is the maximum of the set decyi r 1 e n req i max dec rr 2 j i l dec This required deceleration along with the speed of the vehicle relative to the posted speed would be used to establish the signal intensity assigned to the next barrel ri 11 As outlined previously when decyeg exceeds a minimum threshold the associ
8. eing reflected from an object Variation in the color and texture of the object will change the amount of reflected light and hence the detection range Output light is modulated to differentiate it from external sources such as sunlight Units are available with an output signal proportional to the distance to the object or a logical output that indicates an object is within a specified range This technology consumes more power than the above two and also is much more sensitive to ambient conditions Speed is calculated from the time it takes an object to travel between the two sensor beams 4 4 PULSED DOPPLER ULTRASOUND This solution is different from the side looking configurations already examined because it uses only one sensor mounted on the barrel that is focused towards oncoming traffic as shown in Figure 9 A single piezoelectric transducer is used to transmit series of pulses and then receive the reflected signal Units are available with frequencies in the 30 50 KHz range beam widths from 8 to 30 degrees and ranges to 50 feet 15 Figure 9 Speed measurement with pulsed Doppler ultrasound A microcontroller can calculate range by measuring the time lapse between the outgoing pulses and the reflected pulses Speed can be calculated by measuring the frequency of the returning pulses and subtracting it from the outgoing frequency to yield the Doppler frequency Because the speed of sound in air is a function of temperature the ambien
9. ensive short range devices deployed in substantial numbers at the work zone in the form of smart barrels Such barrels would nominally have the appearance of today s common work zone traffic control barrels Each barrel would communicate with supervisory computers that would process the distributed speed measurements and dispense commands for appropriately adjusted and distributed signal intensity based on existing speed differentials The system is intended to be readily deployable requiring little more effort than currently needed to set up today s systems of traffic control devices at work zones Following the presentation of the system concept the document discusses what is seen as probably the most challenging technical problem of such a system a sufficiently inexpensive but technically capable traffic speed sensor In successive sections the document reviews detector requirements existing detection technologies technologies selected for evaluation and a plan for prototype testing and evaluation This document concentrates on the speed sensing and speed data processing aspects of the system Concepts for the signaling device are discussed in the companion interim report by John Sullivan Work Zone Safety ITS Human Factors Analysis And Pilot Research Proposal UMTRI May 2004 Among the more promising concepts 1s a pair of alternately blinking lights mounted on the barrel where intensity adjustment would be manifest i
10. eralized flow diagram of the data collection and calculation process envisioned for the site supervisor The process is shown as a continuous loop whose cycle time would be on the order of 10 Hz or more As the first step of the process at the top of the figure speed data is collected from each of the distributed speed sensors These data include the speed value the sensor ID providing location and a time stamp Note that if traffic speed and density are relatively high a given sensor may well have made more than one measurement in the preceding cycle period On the other hand other sensors may have had no vehicles pass by during the cycle period in which case they would transfer no new data Following data input each new data point is checked for validity This is primarily a reality check to remove spurious data Each new data point would be compared with current and recent data from neighboring sensors and impossible readings would be discarded Valid data would be added to the data record Input time stamped speed data acquired during this time step Some sensor may provide no new data Test for validity fail Are new data rational relative to Discard neighboring current and recent data pass Test speed data for currency old Is vehicle likely still in zone Discard current recent Queue evaluations where appropriate Calculate and output signal intensities Figure 5 Generalized flow dia
11. f Vehicle Detection and Surveillance Technologies used in Intelligent Transportation Systems The Vehicle Detector Clearinghouse 2000 http www nmsu edu tratfic Minnesota Department of Transportation Office of Traffic Security and Operations and SRF Consulting Group Inc Portable Non Intrusive Traffic Detection System 18 arti Research of sara ronan Kinet ds 2005 SRP No 4869 O Steve J W1 232DTS User s Manual Rev 1 0 Radiotronix Inc 2002 http www radiotronix com datasheets W1 232 User s Manual revi 2 pdf Szewczyk Robert Eric Osterweil Joseph Polastre Michael Hamilton Alan Mainwaring and Deborah Estrin Habitat Monitoring with Sensor Networks Communications of the ACM 47 no 6 2004 34 40 http portal acm org portal cfm 19
12. ffic signaling system for work zones based on a distributed system of traffic speed sensors and traffic signaling devices Both the sensors and the signals are seen primarily as rather simple inexpensive short range devices deployed in substantial numbers at the work zone in the form of smart barrels Such barrels would nominally have the appearance of today s common work zone traffic control barrels Each barrel would communicate with supervisory computers that would process the distributed speed measurements and dispense commands for appropriately adjusted and distributed signaling based on existing speed differentials The system is intended to be readily deployable requiring little more effort than currently needed to set up today s systems of traffic control devices at work zones The document discusses what is seen as probably the most challenging technical problem of such a system a sufficiently inexpensive but technically capable traffic speed sensor In successive sections the document reviews detector requirements existing detection technologies technologies selected for evaluation and a plan for prototype testing and evaluation A bibliography of technical and commercial references on sensor technologies is included Concepts for the signaling devices are discussed in the companion interim report by John Sullivan Work Zone Safety ITS Human Factors Analysis And Pilot Research Proposal UMTRI May 2004 17 Key Words
13. gram of speed data collection and signal intensity calculations The data is then checked for currency That is given the measured speed the distance between sensors and the time elapsed since the measurement is the vehicle which generated the measurement likely still in the measurement zone Where traffic 1s moving slowly or stopped a measurement taken by a given sensor would stay current for a substantial time and many many process cycles Where traffic is moving rapidly a measurement may derive from a vehicle which passed completely through the measurement zone within a single cycle period Data for vehicles still in the zone of measurement would be designated as current in that zone New data would be current in either the zone of measurement or a down stream zone where the vehicle is predicted to be Where multiple data are current for a single zone the maximum or perhaps average speed would be used to represent the zone The data record would also retain at least several recent readings from each sensor zone primarily for use in the validity check Finally if there is no current speed value for the first barrel upstream of the system an expected speed based on the recent history of speeds of entering vehicles would be assigned as the current speed for the first zone The data can then be further processed to evaluate the queue growth where this is appropriate Whether or not this is worthwhile undertaki
14. icles in the zone Using a representative vehicle spacing in the queue the tail of the queue relative to the second sensor can be estimated A modest complication arises from estimating how many of the vehicles in the zone are still in relatively free motion prior to reaching the queue The final step in the processing cycle is the calculation and output of the signal intensities for each zone Figure 6 presents the nomenclature used in the explanation of that calculation that follows which deals with areas of closely spaced sensors and signals The figure shows that there are n 1 smart barrels in the system each here assumed to have a speed sensor and a signaling device and numbered from 0 through n Bo being the first barrel A full matrix of the distances between barrels is known a priori the distance from the barrel i to barrel j where j gt 1 being designated djjj Each barrel also has an associated elevation e speed value s and signal intensity r Assuming close spacing of the smart barrels per the previous discussion determining signal intensity proceeds as follows don diojfn 1 choy 2 djoyrt Distances d lin dh4j n 1 dr 12 di2lfn df2 n 1 Barrels Bo B4 Bo Bnet Bh Elevation ep e e2 Cn 4 en Speed reading So S4 So Sn 1 Sh Signal intensity ro l4 fo h 1 h traffic flow Figure 6 Nomenclature For positions i 0 through n 2 for which the speed readings si are current a matrix of d
15. mal daily field fluctuations A 3 axis magnetic sensor hybrid Honeywell HMC2003 3 permalloy magneto resistive sensors with onboard signal conditioning on each side of the barrel The six signals x1 y1 z1 x2 y2 and z2 will be digitized and the corresponding magnitudes square root of the sum x y z of the magnetic variations will be subtracted as above As above sensor heights and sensitivities will be adjusted to optimize the result The six direction component signal will also be examined in case the speed algorithm needs to be improved A pair of the Banner Engineering M Gage S18m detectors will also be evaluated This detector has only a vehicle detected yes no output and therefore the two sensors 14 will not be connected differentially as above Instead the time between detection activations will be used directly in the speed calculation 4 3 DUAL ACTIVE INFRARED As shown in figure 8 dual presence sensors determine speed by observing the time period between the two signals Distance between sensors Time between signals Figure 8 Speed measurement with dual presence sensors Packaged active red or infrared sensors are available from several companies that target automated assembly or conditioning machines A diffuse reflective sensor consists of an irradiating usually infrared element and an element photosensitive to the irradiated light in the same unit The emitted light reaches the detector after b
16. n blink rate and duty cycle In interpreting this document the reader is advised to simply assume such a signaling device 2 CONCEPT OF THE WORK ZONE SAFETY ITS SYSTEM The work zone safety ITS system presented here is intended to provide distributed speed advisory signaling which automatically adapts to the current traffic flow situation in the work zone As shown in figure 1 the system is primarily based on the notion of a smart barrel a device similar in appearance to today s work zone traffic control barrel but containing a short range traffic speed sensor a simple but adjustable signaling device and short range communication equipment for interfacing with supervisory computers Such smart barrels would be distributed in large numbers and at relatively short intervals throughout the work zone as ordinary traffic control barrels are distributed The distributed traffic speed data would be received and processed by the supervisory computers to provide rapid real time adaptation of the distributed signals as appropriate for the existing speed differential through out the work zone Site Supervisor One Sue Supervisor Per Site Barrel Supervisors N Barrel Supervisors Smart Barrels AI Barrels pee Superviser Warning Zone Hazard Zone Inform drivers of impending hazzard Traffic is slow or stopped Give them enough time to slaw down Figure 1 System concept 2 1 SYSTEM CONFIGURATION The system consists of smar
17. ng 1s largely determined by the spacing of sensors Where sensors are relatively closely spaced on the order of the stopping distance of vehicles at the nominal speed of travel or less queue evaluation is not necessary For example with sensors in every smart barrel and barrels spaced by the often used rule of spacing in feet equal to two times the speed in miles per hour queue evaluation would not be needed That is large differences in speed between adjacent sensors would constitute an un salvageable situation to be effective warnings would have had to be sent and responded to at previous sensor signal locations Where spacing between sensors approaches an order of magnitude greater than stopping distance queue evaluation may be appropriate In this situation large differences in velocity may exist between adjacent sensor positions that could be manageable and might benefit from adjustment of signal intensity depending on estimates of end end position between the two sensors On the other hand it can be reasonably argued that since the signaling device located at the lead sensor represents the last opportunity to warn of the forthcoming queue the intensity of this signal should be high regardless of the distance to the queue Queue evaluation is a conceptually simple process wherein vehicles entering the zone 1 e passing the lead sensor and leaving the zone passing the second sensor are counted in order to keep track of the number of veh
18. of truth data we will also drive UMTRI vehicles with their own DAS system though the test site on multiple occasions during the field test Data collected on the vehicle will include GPS position and time as well as the obvious vehicle speed GPS data will allow reliable matching of speed measured on the vehicle with speed measurements at the field test sites Of course video taken at the site and synchronized with site data provides and additional cross check In our original proposal we proposed that sites for field testing would in fact be work zones But on further reflection and as implied by this example it is not clear that this is necessary nor perhaps even advantageous 17 6 BIBLIOGRAPHY Banner Engineering Corp M Gage S18m Vehicle Detection Sensor Data and Application Sheets 2004 www bannerengineering com Caruso Michael J and Lucky S Withanawasam Vehicle Detection and Compass Applications using AMR Magnetic Sensors Honeywell SSEC 1999 www ssec honeywell com CTC amp Associates LLC WisDOT RD amp T Program Evaluating New Sensor Technologies for Actuated Signals Report prepared for Bureau of Highway Operations Division of Transportation Infrastructure Development December 2003 http www dot wisconsin gov library research docs tsrs tsrsignalsensors pdf Culler David E and Wei Hong Wireless Sensor Networks Communications of the ACM 47 no 6 2004 30 33 http portal acm
19. of the experimental nature of the system development and ease of programming A final product might use a simple microcontroller The system contains one or two of the same wireless modules that are in the smart barrels In addition it needs a higher power longer range radio modem to talk to the site supervisor The higher power consumption of this system will probably require a solar panel to charge the battery Small Antenna Small Antenna Antenna Lights Figure 3 Configuration of a barrel supervisor The barrel supervisor collects speed readings from the barrels and forwards them to the site supervisor It also receives signal commands and time synchronization messages from the site supervisor and relays those to its barrels It would also participate in setting the barrels to a low power standby state and then reawakening them when traffic approaches Site Supervisor The site supervisor is the brains of the system A possible hardware configuration is shown in Figure 4 It has a GPS receiver to synchronize its clock and to locate it and the smart barrels A GPRS cellular modem is provided for reporting to funding or managing agencies Since all of the barrels have Lat Lon coordinates established upon setup automatic reporting of work zone location extent and traffic conditions is possible Antenna Figure 4 Configuration of the site supervisor The site supervisor receives time stamped ID ed speed reading
20. org portal cfm Goode Barbara G ed Wireless for Industry Supplement to Sensors Magazine Summer Hill Jason Mike Horton Ralph Kling and Laksman Krishnamurthy The Platforms Enabling Wireless Sensor Networks Communications of the ACM 47 no 6 2004 41 46 http portal acm org portal cfm Klein Lawrence A Sensor Technologies and Data Requirements for ITS Boston Artech House 2001 http www artechhouse com Martin Dr Peter Yugi Feng Xiaodong Wang Detector Technology Evaluation Department of Civil and Environmental Engineering University of Utah Traffic Lab 2003 http Awww tratficlab utah edu Middleton Dan Deepak Gopalakrishna and Mala Raman Advances in Traffic Data Collection and Management One of three white papers that support a series of workshops on Data Quality which were held in March 2003 The EDL numbers for the full set of white papers are 13766 13767 13768 Prepared for Office of Policy Federal Highway Administration Washington DC 2003 http www itsdocs thwa dot gov JPODOCS REPTS TE 13766 html Middleton Dan Ph D P E and Rick Parker Vehicle Detection Workshop Participant Notebook Texas Transportation Institute The Texas A amp M University System 2000 http rce tamu edu docs RCE PDF Vol I Tamu Middleton TMO amp pdf Mimbela Luz Elena Y Lawrence A Klein Ph D P E Perry Kent John L Hamrick Karen M Luces and Sylvia Herrera A Summary o
21. r consumption 20 50 watt range make them unsuitable for this application The dirty construction environment and changing road features might also prevent their use 13 4 SELECTED TECHNOLOGIES FOR EVALUATION After accurately measuring vehicle speed low power consumption is the next most desirable feature of a speed sensor technology that will run off of batteries UMTRI will evaluate the sensors from lowest to highest power usage 1 e passive infrared magnetic active infrared and ultrasonic 4 1 DUAL PASSIVE INFRARED An infrared detector consists of a lens a filter opaque to visible light and a sensing element Figure 7 illustrates how a dual sensor measures speed The difference of the sensors is the quantity of interest Infrared sources seen by both sensors are removed from the signal As a vehicle passes the barrel the heat is detected first in sensor 1 and then in Sensor 2 The sensor separation 18 inches divided by the time between the peaks equals the vehicle speed Sensor height lens field of view and gain sensitivity will be varied to attempt to sense only targets in the immediate adjacent lane f 18 inches apart Barrel width Figure 7 Differential Sensor Configuration 4 2 DUAL 3 AXIS MAGNETOMETER The dual 3 axis magnetometer will be configured similar to the infrared solution depicted above This makes it easier to cancel out the magnetic field changes caused by local metal objects and the nor
22. s from the barrels via the barrel supervisors The algorithms described in the next section create signaling commands that are sent out to the barrel supervisors and then to the barrels or other connected driver warning or communication devices 2 2 SETUP AND CALIBRATION Barrels are turned on and placed along the lane s of travel with attention paid to proper alignment relative to the traffic lane of interest When a set of barrels has been placed a configuration box is used to send an initialization message to the barrels to tell them to switch their radios to the proper channel and then to enter a low power active standby mode After the barrel supervisors and site supervisors are powered and initialized the barrels are switched to active mode and begin speed sampling Next the barrels need to be located A laptop computer with a 10 or 20 Hz differential GPS is placed in a suitable vehicle The vehicle is driven three times in each lane of interest over the length of the deployment GPS time altitude latitude and longitude are collected and written to file or files on the laptop hard disk The files are then transferred to the site supervisor computer via an ethernet connection The site computer imports these files and uses the times in the file to connect the vehicle path to the time stamped speed data collected from the barrels Thus the barrels are located on the map Note that more precisely it is the position of the vehicle at
23. section 5 we will evaluate magnetometers mounted on plastic barrels to sense velocity 12 3 6 MICROWAVE Continuous microwave devices use the Doppler principle the change of frequency of a wave reflected from a moving object is proportional to the objects speed to directly measure the speed of a vehicle The available systems TDN 30 by Whelen Engineering TC 20 by Microwave Sensors and DRS1000 by GMH Engineering are configured to be used in an overhead mount or an elevated side mount Typical power consumption is 2 to 6 watts and single unit cost is 1000 to 2000 It might be possible to mount a radar device on top of a barrel facing towards oncoming traffic The high cost high power consumption and possible interference problems with multiple radars in the same location would make this only a fallback solution 3 7 VIDEO Video image processing promises the richest data set of traffic measurements including vehicle detection speed classification headway density and volume A camera sends an image to a video processor which digitizes it and applies various detection and tracking algorithms to extract the applicable measurements Cameras can be side or overhead mounted The Autoscope Solo Econolite Control Products even integrates the camera and processor into one package Video systems are probably the hardest to install and calibrate correctly of all detector technologies The lack of portability high cost and the high powe
24. supervisor with an incorporated signaling device The calculation of signal intensity would however proceed on the same basis as indicated above perhaps with the addition of queue evaluation in the intervening zone to the next barrel 10 3 REVIEW OF DETECTION TECHNOLOGIES PRODUCTS Most of the existing traffic detection and surveillance products are targeted to either signalized intersections or freeway applications The intersection products detect vehicle presence and are used to activate traffic signals Inductive loops buried in each lane are the most common detector Direct speed sensing 1s not the primary measurement Freeway applications emphasize vehicle counting and classification and usually report average speeds The non intrusive sensors are usually mounted on existing structures signs bridges mast arms and poles and get their power from 110 or 230 volt connections to the power grid or use DC power supplied by traffic control cabinets The power consumptions range from 1 to 160 watts Large size restrictive mounting requirements or excessive power consumption preclude using many of the existing products examined below Some of the sensor technology however can be adapted to be used in this application 3 1 ACTIVE INFRARED An active infrared sensor sends out infrared light generated by a LED or laser diode and measures the time required to reflect off an object and return to an infrared detector or array of detec
25. t barrels barrel supervisors and a site supervisor Smart Barrel Figure 2 shows the major elements that comprise a smart barrel A low power microcontroller contains a local clock that is synchronized to GPS time via inputs from the barrel supervisor It has program logic and hardware interface elements to derive speed from the detector s and control the on off state of the signals LEDs It should also be able to go into a low power sleep mode when low traffic density permits A message containing the barrel identification time and vehicle speed is sent when a vehicle or portion of a vehicle is first detected Closely spaced vehicles may not require a new communication Figure 2 Configuration of a smart barrel Individual barrels will not include GPS hardware However each barrel s LAT LON will be determined during set up See section 2 2 A tilt sensor will be contained in each barrel to warn the supervisor if the barrel has moved since setup Each barrel would be powered by its own battery providing long term operation on a single charge Low power communications ability is essential for this to succeed For purposes of explanation an actual radio module will be described here even though a fully communicating system may be beyond the scope of this phase of the project The Wi 232DTS Radiotronix embedded wireless module combines a high performance DTS spread spectrum transceiver and a protocol controller to create a
26. t temperature must be measured and used to correct ranges The HE661 Hexamite signal conditioner interfaces with the company s line of ultrasonic sensors and outputs the range to the closest object and the speed of the object traveling at the maximum speed within the operating boundaries 4 5 DUAL ULTRASONIC PRESENCE Finally two pulsed ultrasonic sensors at lower power outputs than the above can be mounted in the same configuration as the active infrared solution Here each would detect only vehicle presence and the speed would be calculated from measuring the time it takes to travel between detectors Possible interference between sensors might require some electronically controlled phasing or a mechanical separation although the barrel could be enough 16 5 SENSOR EVALUATION AND TESTING UMTRI intends to evaluate and test the selected sensor technologies in a two stage process made up of 1 evaluations of the selected technologies in controlled testing situations until two or three are deemed acceptable and 2 a limited field test of the more promising sensors identified by the evaluation UMTRI will purchase or assemble one speed sensor of each of the five selected technologies Each sensor will be evaluated in a set of controlled experiments until two or three are found worthy of further testing Data will be collected using UMTRI s existing Data Acquisition System DAS Target vehicles will be UMTRI owned vehicles driven p
27. tensity would be adjusted either proportionately or in steps to progressively higher levels up to a maximum associated with upper deceleration threshold Although the specific thresholds of deceleration remain a matter of study in the simulator and would at any rate be adjustable in a final system our expectation would be for a minimum threshold of about 0 05 g 0 5 m sec and an upper threshold of about 0 15 g 1 5 m s2 This expectation derives from recognizing that 1 the deceleration capability of heavy trucks not passenger cars is the more important reference and 2 our understanding of the distribution of braking deceleration of trucks in real use It also should be noted that in practice deceleration thresholds would be adjusted for the grade That is on descending road segments the thresholds would be reduced in accordance with the downward grade and similarly increased on ascending road segments A similar adjustment of intensity would be based on over speeding relative to the posted speed limit In a fashion analogous to the deceleration procedure this calculation would define a minimum threshold of over speed and an upper threshold and vary intensity in proportionately or step wise across the range they define Finally of course the actual intensity setting used would be the higher of the respective results of the deceleration and the over speed criterion Data collection and processing Figure 5 presents a highly gen
28. the moment the barrel senses its speed that is actually determined This in fact is the more desirable location information Distances and grades between barrels is then calculated and stored for use in system operation see section 2 3 The site supervisor can now start its normal operating algorithms 2 3 OPERATION Determining signal intensity based on distributed speed measurements Actual development of full system software is beyond the scope of this phase of the Work Zone Safety ITS project Nevertheless it is necessary to develop system algorithms at least at the conceptual level in order to have a basis for design of the hardware elements discussed above The following discussion presents the structure of the core algorithms at this basic level For the moment these concepts are limited to single lane traffic although we believe they will be readily adaptable to multilane situations Signal intensities presumably to be manifest as blink rates to be displayed at each individual signaling device are expected to be set primarily on the basis of the deceleration required of faster moving traffic to avoid collision with slower moving traffic ahead but with additional adjustment based on over speeding relative to the posted speed limit Addressing the deceleration basis first signals would not be activated unless the required deceleration exceeded a minimum threshold When required deceleration did exceed the minimum signal in
29. tors The Autosense Scwartz Electroptics series of sensors scan across multiple lanes and can provide 3D imagery to classify vehicles Speed can also be calculated by measuring the time it takes the vehicle to cross detection zones The Traffic observation module MBB Sens Tech does not scan but uses two to six laser beams to create several detection zones These sensors mount 20 to 25 ft above the road They are very accurate but much higher in cost 5000 10000 and energy consumption 40 to 160 watts than the other technologies A scaled down solution that uses infrared emitters and detectors in a side looking configuration is discussed in section 4 3 2 PASSIVE INFRARED All matter above absolute zero emits radiation in the far infrared part of the spectrum The amount of radiation is a function of the object s temperature size and structure Passive infrared sensors measure this radiation A non imaging detector has a wide field of view and can detect a vehicle s presence or velocity with more than one sensor or detection zone An imaging sensor contains a two dimensional array of detectors and so can indicate presence speed and classification The IR 254 ASIM Technologies Ltd has four detection zones mounts overhead to 33 ft costs 955 and consumes just 0 3 watts Passive infrared is a promising technology for this application because of its low power consumption cost and an adaptability to a side viewing detector Many sec
30. transparent wireless solution to replace conventional RS 232 422 485 wiring It can be used in point to point point to multipoint or multipoint to multipoint applications Its footprint is less than one square inch and cost is estimated at 20 in production quantities The module is designed to interface directly with standard UART signals from a microcontroller It can use a printed circuit board antenna or an external 1 4 wave whip antenna In this application the set of barrels attached to a barrel supervisor would share a single data channel and part of the maximum data rate of approximately 150 K Baud The set of barrels and their supervisor should be in line of sight within a maximum range of 1000 to 1500 feet Thus the number of barrel supervisors is determined by barrel spacing and the terrain The detectors should reliably sense the speed of vehicles in the immediately adjacent lane Stopped vehicles are not directly transduced but are deduced see section 2 3 Detectors should also consume as little power as possible and be immune to variations in weather vibration and other disruptive influences as found in a typical roadside application Barrel Supervisor The barrel supervisor acts as a gateway between its barrels and the site supervisor It could be attached to the back of a sign that is also used to signal drivers The main parts are diagrammed in Figure 3 The heart of this system is a low powered embedded PC chosen because
31. urity products use infrared motion detectors A speed sensor is described later in section 4 1 11 3 3 PASSIVE ACOUSTIC A passive acoustic sensor detects sound primarily tire noise from approaching vehicles with a two dimensional array of microphones The array is mounted on a pole 20 to 40 feet above the ground beside the road SmartSonic and SAS 1 sensors are two passive acoustic products They both have trouble with slow moving vehicles and stop and go traffic They are primarily recommended for measuring free flow traffic at speeds above 30 mph and therefore are inappropriate for a work zone application 3 4 ULTRASONIC Ultrasonic air sensors emit a burst of sound pulses at a frequency between 25 and 50 KHz Distance is calculated by measuring the time it takes the beam to return after reflecting off the target Two closely spaced emitters can permit speed measurement Sensors can be overhead or side mounted The TC 30C Microwave Sensors is a an ultrasonic ranging sensor that indicates vehicle presence It costs 559 and consumes 3 to 4 watts There are dozens of ultrasonic emitter transducer and integrated sensor manufacturers whose products are used in object detection in commercial and industrial applications and are fairly inexpensive An ultrasonic detector that also measures the Doppler frequency shift of a reflected signal is available for purchase The two main problems with these types of detectors are high power consumption
Download Pdf Manuals
Related Search